RX 7900 XTX
6recorded runs
550.2best tok/s
6sourced
How it compares — 24 GB tier
What fits on 24 GB — measured
| Model | Quant | Peak VRAM | Headroom |
|---|---|---|---|
| Qwen3.8-27B | Q4_K_M | 22.9 GB | 1.1 GB |
| Qwen3.6-35B-A3B-abliterated-v4 | Q8_0 | 21.2 GB | 2.8 GB |
| Qwen3.8-27B | MQ4 | 17.1 GB | 6.9 GB |
All benchmarks Compare RX 7900 XTX →
| GPU | Model | Quant | Backend | Ctx | tok/s | prompt tok/s | VRAM GB | Source |
|---|---|---|---|---|---|---|---|---|
| RX 7900 XTX · 24 GB | Qwen3.5-0.8B | MQ4 | hipfire 0.3.0+44eb828cc804 | 128 | 550.2 ✓ | — | — | src |
| RX 7900 XTX · 24 GB | LFM2.5-350M | MQ4 | hipfire 0.3.0+b0bcc3f91506 | 4096 | 398.4 ✓ | — | — | src |
| RX 7900 XTX · 24 GB | Qwen3.8-27B | MQ4 | hipfire 80a572c | 2048 | 154.2 ✓ | 51.3 | 17.1 | src |
| RX 7900 XTX · 24 GB | Qwen3.6-35B-A3B | MQ4R | hipfire 0.3.0+32d27a87d7f0 | 4096 | 150.0 ✓ | — | — | src |
| RX 7900 XTX · 24 GB | Qwen3.6-35B-A3B-abliterated-v4 | Q8_0 | llama.cpp b9870 | 262144 | 82.1 ✓ | 2123.1 | 21.2 | src |
| RX 7900 XTX · 24 GB | Qwen3.8-27B | Q4_K_M | llama.cpp b10791 | 2052 | 58.0 ✓ | 921.0 | 22.9 | src |